Transcript
This podcast comes to you from the frontiers of electronics research. It is the first in a series from Surrey University, developed and presented by Dr Radu Sporea, looking at the latest developments in electronic engineering – technologies that still have a long way to go before they become part of everyday products, but which show real promise of changing the way we think about energy efficient electronics.
The integrated circuit has been serving us well for 50 years, and over that time silicon has been the element of choice for electronics. But how silicon is being used is changing, and Radu and his colleagues at Surrey University, Dr Charles Opoku and Stamatis Georgakopoulos, are looking at the potential of other materials, too, sometimes with surprising results..
Dr Radu Sporea: If you were listening to this podcast during Christmas shopping season, there would be a high chance that you would have some sort of electronic gadget on your shopping list. Every year we expect without even thinking about it to be able to buy faster computers, smartphones and electronics cheaper than the previous year’s models. This relentless evolution of technology has been down largely to the clever improvements in the manufacturing of electronic circuits. These days, the circuits have billions of microscopic components, and that complexity is possible in the first place because we make all these parts from a single piece of material. For this reason speed and reliability are ever higher, while cost is minimized. We call this invention the integrated circuit, and while the concept is not new, it’s been serving us well for nearly half a century. Just as the integrated circuit was flourishing in the 1960s, Gordon Moore, a lead technologist at Intel, observed that if every year engineers made circuits with components half as small and twice as many, within a very short time, extremely complex circuits would be made but cost would stay about the same. Moore’s law, as we’ve come to call it, has become a self-fulfilling prophecy, serving as a guideline for the future for companies involved in electronic circuit manufacturing.
Digital circuits are fabricated in semiconductors, materials which can change from being good conductors of electricity to being insulators. Circuit designers rely on manipulating this property of semiconductors in order to store data or make computations. And for nearly as long as integrated circuit have existed, one semiconductor has been the material of choice for the overwhelming majority of electronics: silicon. Dr. Charles Opoku explains why we prefer it above all others:
Charles OPOKU
- Abundant, sand, silicon oxide, so relatively cheap;
- Melt it – ingot – extremely pure – 99.9999…%
- As it cools it becomes a single crystal – atoms are very ordered, very few imperfections so we can make complex, reliable and fast circuits out of it.
- Cut it – thin discs – wafers as large as dinner plates or even bigger
- Many circuits in an array, cut those to bits, stick into a package and we’re done
Radu Sporea: Silicon offers the best balance between performance and cost. Over the decades we’ve been very inventive when it comes to squeezing the last drop of speed or energy efficiency form our silicon circuits. We’ve put the silicon under strain which makes it easier for electrons to go through the material at high speed, and we’ve implanted precise quantities of other chemical elements at specific locations in an attempt to improve their efficiency.
Silicon is cost-effective for circuits with millions or billions of components packed together into the size of a fingernail. But there are integrated circuits which just have to be a lot bigger. Think about your television, or a tablet computer. Their flat screen is a single circuit, made in one piece. It is still an integrated circuit – we call it a large area electronic device, but making it on gigantic silicon wafers would be prohibitively expensive and quite frankly pointless. The number of components in the whole screen is a few million, but not anywhere near as many per square inch as the most advanced digital circuits, so using the same technology would mean over engineering. In fact, any circuit roughly bigger than a two-pound coin becomes too expensive to make on a silicon wafer. A cheap way of fabricating large-area electronics is needed.
So, on the glass support that eventually become the television screen, engineers deposit a layer of Silicon a thousand times thinner than a sheet of paper. They use it to make the electronic circuits that control the colour and brightness of every part of the display screen. But, as PhD student Stamatis Georgakopoulos explains, putting silicon on glass is not without its drawbacks.
Stamatis Georgakopoulos:
- Silicon wafer – extremely low defect density
- Crystal – ordered three dimensional array of atoms
- Electrons which make up the electrical charges flow unimpeded through the atomic array and speed is high
- On glass – when we put down silicon atoms, they arrange at random so rather than an ordered crystalline structure, they create an amorphous layer, in which there is no order
- This makes it hard for electrical charges to travel through the disordered material.
- So hard in fact that their speed and that of electronic circuits made this way drops by a factor of around 1000!
Radu Sporea: Even with such a large speed penalty we still manage to use amorphous silicon in display screens and very successfully: for Smartphones, tablets and HDTVs. But as screen size and image quality increase, so do the demands on the electronics and soon there will come a time when amorphous silicon simply will not be able to rise to the challenge. So what can we do about it?
One option is to attempt to convert the amorphous, disordered silicon into the ordered, crystalline version we know works much faster. To do that, as the material is laid down during fabrication, we use lasers to temporarily heat small areas of silicon to very high temperatures, above the melting point. As it cools down, the material naturally forms small crystals in which we can make our high-speed electronic devices. The trouble is that these crystals tend to vary in their size and properties, so getting consistent performance, say uniform colour from a large flat screen made this way, is quite a challenge.
So how else can we solve the problem of slow amorphous silicon? Well, we could replace it with something entirely different. Charles Opoku:
Charles OPOKU
- A very exciting way of doing it is with Silicon nanowires
- Tiny rods, 1000 times smaller than the human hair
- Each formed of a single crystal of silicon, their structure is very ordered, so their electrical performance is very high, but you have the advantage that you can spread them around with ease and get the benefit of crystalline silicon on a large surface.
- The trick is that not all wires are the same thickness and the way they overlap is random, so large variations in performance from place to place are possible.
- A lot yet to research and to understand, not least which is the best way of using them to minimize these problems.
Stamatis Georgakopoulos
- Yet another solution is to replace silicon altogether.
- Metallic oxides are a class of materials more than ten times more conductive, so in theory you can get a lot faster electronics, but so far they are still being researched.
- They are prone to large performance variations if exposed to moisture or oxygen from the atmosphere
Radu Sporea: Some manufacturers are already showing prototypes of high performance display screens made with zinc oxide, while silicon nanowires are still in the fundamental research phase. Their chemical composition and geometrical properties make them behave quite differently from conventional flat layers. To make real products with these materials we first need to understand the physics which govern the movement of electrical charge within them. But equally important is studying the ways in which these materials need to be fabricated and used to benefit the most from their specific properties.
Charles Opoku
- There are always surprises when fabricating new devices or working with new materials.
- With nanowires, you think you are doing something but something unexpected happens because these materials behave differently that what you’re used to.
Stamatis Georgakopoulos
- Then you try to understand why things are happening the way they are – only then you may be able to really make good use of these materials and exploit their particular properties to create something which was not possible before.
- You need a lot of patience and flexibility in the way you think.
Radu Sporea: As we will see in the next episode, a flexible approach may lead to surprising new ways of doing things. Watch this space on Pod Academy!
For more on this and other aspects of electronic engineering research see some of Radu Sporea’s YouTube videos:
http://www.youtube.com/playlist?list=PLF5CFD031EC74DDE1&feature=mh_lolz
http://www.youtube.com/watch?v=SnypTySMv_o
http://www.youtube.com/watch?v=xv-z9VjR50U
Tags: Electronics
I was curious if you ever considered changing the structure of your blog?
Its very well written; I love what youve got to say.
But maybe you could a little more in the way of content so people could connect
with it better. Youve got an awful lot of text for only having
1 or 2 pictures. Maybe you could space it out better?
Very interesting and helpful information. Bookmarked !
Appreciate the great read.